4 Reaction Mechanisms Part (iii) Free-radical Reactions By D. GRILLER Nationai Research Council of Canada Ottawa Ontario Canada KIA OR6 1 Synthesis A new method has been discovered for the reductive dehalogenation of aryl vinyl cyclopropyl and bridgehead iodides.' The approach involved photolysis of mixtures containing the halide di-t-butyl peroxide and lithium aluminium hydride in THF; yields (g.c. analysis) were normally ca. 90%. The chain propagation steps are shown in Scheme 1. When Ar was neophyl substantial amounts of product were derived from the rearranged form of the radical suggesting that the second step of the reaction was relatively slow (rate constant <lo3dm3 mol-' s-'). ArX + AlH, -+ Ar-+ AlH3X-Ar. + AIH; -ArH + AIH35 Scheme 1 Barton et al.devised a high-yield 'one pot' method for the decarboxylation of acids.2 The 0-ester of N-hydroxypyridine-2-thionewas prepared and was then reduced by tri-n-butyltin hydride or t-butyl mercaptan giving >70% yield of isolated product (Scheme 2). However when R was tertiary the mechanism involved forma- tion of the corresponding pyridyl sulphide which was then reduced by the tin hydride to the desired product (Scheme 2). Several reports have dealt with the use of radical cyclization reactions in chemical synthesi~.~~ In the most intriguing of these the cyclized radical was scavenged by t-butyl isocyanide. Subsequent elimination of t-butyl left the chemically versatile cyano-group neatly incorporated in the product (Scheme 3). Photolysis of hexaphenylditin was used to generate the stannyl radicals needed for the initial bromine abstraction.Tin hydrides could not be used since they proved to be far more effective as radical scavengers than the isocyanide. ' A. L. J. Beckwith and S. H. Goh 1.Chem. SOC.,Chem. Commun. 1983 907. ' D. H. R. Barton D. Crich and W. B. Motherwell J. Chem. SOC.,Chem. Commun. 1983 939. E. J. Corey G. Schmidt and K. Shimoji Tetrahedron Lett. 1983 24 3169. G. Stork and R. Mook jun. J. Am. Chem. SOC.,1983 105 3720. G. Stork R. Mook jun. S. A. Biller and S. D. Rychnovsky J. Am. Chem. SOC.,1983 105 3741. G. Stork and P. M. Sher J. Am. Chem. SOC.,1983 105 6765. 87 D. Griller S (1) + Bu",Sn. -+ R. + C02 + SSnBu R-+ Bu",SnH 3 RH + Bu",Sn* For R = tertiary alkyl SR Scheme 2 OEt c!AJBr Ph,Sn- -• OEt OEt Scheme 3 The synthesis of eight trans-alkyl hyponitrites RO-N=N-OR has been repor- ted.' They were generally prepared by the addition of silver hyponitrite to the corresponding alkyl iodide or bromide at 0°C in pentane as solvent.Most were white crystalline solids which could be stored for long periods without change below ' C. A. Ogle S. W. Martin M. P. Dziobak M. W. Urban and G. D. Mendenhall J. Org. Chem. 1983 48 3728. Reaction Mechanisms -Part (iii) Free-radical Reactions 0 "C. The hyponitrites decompdsed at 66 "C in iso-octane solvent with first-order kinetics and with half-lives (min) ranging from 32.3 (R = Me) to 5.5 (R = 1-phenylethyl). They should prove to be useful low-temperature thermal initiators.2 Mechanism 1983's most fascinating and controversial papers on mechanism dealt with the possibility that radicals can be generated thermally in two distinct electronic states. Skell and May investigated the reactions of acyloxyl radicals at -78 "C in mixtures containing methylene chloride and neopentane.' They found that in addition to the normal Hunsdiecker reaction a small amount (ca. 10%) of hydrogen abstraction was also taking place. The abstracting agent was taken to be the acyloxyl radical on the basis of the relative reactivities displayed by the two substrates which were reported to be quite different to those expected if alkyl radicals or halogen atoms had been the abstracting agents. However product yields for the carboxylic acids were thought to be unreliable and were not investigated in detail (Scheme 4).Hunsdiecker RCO,. -+ R-+ C02 Re + RC0,Br -+ RBr + RCO,. Abstraction RC02. + R'H + RCO2H + R'. R'. + RC0,Br --* R'Br + RC02* Scheme 4 The experiments were carried out first in the presence of bromine and then in the presence of vinylidene chloride which was used as a bromine atom scavenger. The relative reactivities of the substrates differed by a factor of ca. 2 under these conditions and it was therefore proposed that the acyloxyl radical could be generated in two electronic states (Scheme 5). With bromine present R'. + Br -+ R'Br + Br-Br. + RC0,Br -+ Br + RCO,. 'lT' With a bromine scavenger R'. + RCOzBr -+ R'Br + RC024 'IT' Scheme 5 Numerous control experiments were carried out in an attempt to show that the presence of other chain carriers was not responsible for the effect.Moreover when 1-bromobutane isobutane and n-butane were used as substrates the selectivities P. S. Skell and D. D. May J. Am. Chem. SOC.,1983 105 3999. 90 D. Griller for the proposed u and T acyloxyls continued to show small but significant differen- ces which overall were quite different to those for bromine atoms and alkyl radicals. Almost identical experiments and arguments were advanced to support the inter- mediacy of (T and T succinimidyl radicals.' These have now been challenged and two groups have argued against the concept.lO," One group accounts for the changing patterns of selectivity in terms of a bromine atom chain a succinimidyl chain or a mixture of the two depending upon the experimental conditions." The second favours as possible chain carriers the bromine atom the succinimidyl radical and a bromine atom-N-bromosuccinimide complex.' I However it is already clear that some of the experimental evidence involved in one of the challenges'' is suspect.' However the literature contains an additional counter argumentI2 to the u,T concept which has yet to be addressed by its proponents.Baban Brand and Roberts have shown that borane radical anions can be gener- ated by photo-induced electron transfer from hydroborate anions.I3 The reaction was only successful when the light was absorbed by the charge-transfer-to-solvent band of the anion in solution.Thus photolysis of Bu",NBH in liquid ammonia-1,2- dimethoxyethane (A < 290 nm) gave the e.s.r. spectra of H3BT and the solvated electron which had lifetimes of ca. 108 and 134 ms at -23 "C.When t-butyl chloride was added to the reaction mixture these spectra were replaced by that of t-butyl (Scheme 6). A typographically corrected paper on borane radical anions was also p~b1ished.l~ H,B.&, + Bu'Cl -+ Bu'. + H3BCl-e& + Bu'Cl -+ But. + C1-Scheme 6 Product studies on the fate of conformationally biased cyclohexyl radicals such as 4-t-butyl- 4-t-butyl-cis,czs-2,6-dimethyl- and 4-t-butyl-cis,trans-2,6-dimethyl-cyclohexyl showed that loss of hydrogen to form olefin was favoured when the C-H bond involved lay close to the plane of the adjacent semi-occupied 0rbita1.l~ Quinga and MendenhallI6 investigated chemiluminescence intensities during the thermolysis of alkyl hyponitrites containing a-hydrogens.The data were consistent with the mechanism of Scheme 7 where substantial recombination between triplet RON=NOR 4 [RO. t + N + RO. 41 -+ ROH + R'R~CO (Os) 11 fast [RO. t + N2 + RO. t ] -+ ROH + R'R2C0 (3T) Scheme 7 ' P. S. Skell R. L. Tlumak and S. Seshadri J. Am Chem. Soc. 1983 105 5125 and references cited therein. 10 D. D. Tanner T. C.3. Ruo H. Takiguchi A. Guillaume D. W. Reed B. P. Setiloane S. L. Tan and C. P. Meintzer J. Org. Chem. 1983 48 2743. I' C. Walling G. M. El-Taliawi and C. Zhao J. Am. Chem. Soc. 1983 105 51 19. '' A. G. Davies B.P. Roberts and J. M. Smith J. Chem. Soc. Perkin Trans. 2 1973 2221. I3 J. A. Baban J. C. Brand and B. P. Roberts J. Chem. Soc. Chem. Commun. 1983 315. l4 J. R. M. Giles and B. P. Roberts J. Chem. Soc. Perkin Trans. 2 1983 743. l5 A. L. J. Beckwith and C. J. Easton J. Chem. Soc. Perkin Trans. 2 1983 661. '' E. M. Y. Quinga and G. D. Mendenhall J. Am. Chem. Soc. 1983 105 6520. Reaction Mechanisms -Part ( iii) Free-radical Reactions radical pairs took place within the solvent cage. The yield of the triplet pathway was proportional to the exothermicities of the reactions calculated for the ground-state products and not to those of the triplet products. It was estimated that the quantum yield would be as high as 0.5 if the exothermicity of the reaction were ca.385 kJ mol-’. This is probably the first unequivocal demonstration that an appropri- ately chosen triplet radical pair can undergo a disproportionation reaction. The selectivity of chlorine atoms in their reaction with 2,3-dimethylbutane was investigated in benzene-trichlorofluoromethane mixture^.'^ Previous investigations showed that benzene moderated the reactivity of the chlorine atom and a .n-complex had been proposed to rationalize the results. In the more recent work,” it was shown that the selectivity depended upon the concentration of 2,3-dimethylbutane. This should not have been the case if the wcomplex were in rapid equilibrium with free chlorine atoms and benzene and the presence of another intermediate namely a cyclohexadienyl radical was invoked (Scheme 8).However it would seem that the data could be consistent with the irreversible formation of a single intermediate be it cyclohexadienyl or .n-complex. c1 CIS + R* + HCI +C,H Scheme 8 Details of a I3C CIDNP study of arylethyl radicals have been reported.” In the case of the 2-phenylethyl radical no polarization transfer could be found that was consistent with its rearrangement to the corresponding spirocycloalkylcyc- lohexadienyl radical (Scheme 9). However if evidence for the rearrangement is to be obtained from polarization transfer then ring closure would have had to take place before relaxation of the nuclear spin states of the radical (ca. s). A kinetic e.s.r. study” indicates that the rate constant for the rearrangement to the spirocyc- loalkylcyclohexadienyl radical must be several orders of magnitude lower than this and hence the absence of polarization was entirely to be expected.Scheme 9 ” P. S. Skell H. N. Baxter 111 and C. K. Taylor J. Am. Chem. SOC.,1983 105 120. Is G. A. Olah V. V. Krishnamurthy B. P. Singh and P. S. Iyer J. Org. Chem. 1983 48 955 5414. A. Effio D. Griller K. U. Ingold J. C. Scaiano and S. J. Sheng J. Am. Chem. SOC.,1980 102 6063. D. Griller Whereas vicinal 1,2-shifts are actually quite common for carbon-centred radicals they are scarcely known for silicon-centred radicals. Wilt and Keller found that the vicinal migration of an acetoxy-group can take place for a silicon-centred radical.20 Thus photolysis of a mixture containing di-t-butyl peroxide and (acetoxy-methy1)dimethylsilane in hexadecane solvent gave radicals (2) and (3) which were trapped by carbon tetrachloride (Scheme 10).It is clear from the experimental conditions and product yields that the rearrangement in this instance must be far more rapid than that for the carbon analogue which presumably reflects the thermochemistry involved. (2)/(3) + CCl -+ (2)C1/(3)C1 + CCI Scheme 10 Sonication of argon-saturated aqueous solutions was found to give .OH and H. which appeared to have been formed in the cavitation bubbles2’ Evidence for the formation of these species was obtained by using a variety of spin traps. The evidence was reinforced with careful control experiments. For example it was shown that the concentrations of the spin-adducts grew from the start of sonication to steady- state levels and that either or .OH could be diverted by reaction with specific Ha scavengers.Obviously the technique could develop into a very useful method for initiating radical processes in water. 1-Adamantyl and 1-bicyclo[2.2.2]octyl radicals were generated by photolysis of the corresponding azo-compounds in various hydrocarbon solvents.22 Remarkably high isotope effects (kH/kD = 25; kH/k = 104) were found for hydrogen abstrac- tion at cyclohexane by 1-adamantyl. The result was rationalized in terms of a contribution from quantum mechanical tunnelling brought about because of a ‘thin’ activation barrier for the reaction. This was thought to have arisen because very little molecular motion was required to convert the I-adamantyl radical into its parent hydrocarbon.Cyclohexyl and ally1 radicals were detected by e.s.r. spectroscopy when cyclo- hexane and 2,5-dimethylhex-3-ene were deposited respectively on a clean surface of sodium chloride at -196 0C.23 The precise details of the reaction mechanism were unknown but the very high electrostatic fields on the crystal surfaces were thought to play a critical role. In the experiment layers derived from hot sodium chloride vapour and from the organic substrate were deposited alternately on a cold rotating 20 J. W. Wilt and S. M. Keller J. Am. Chem. Soc. 1983 105 1395. 21 K. Makino M. M. Mossoba and P. Riesz J. fhys. Chem. 1983 87 1369. 22 P. S. Engel W.-K. Chae S.A. Baughman G. E. Marschke E. S. Lewis J. W. Timberlake. and A. E. Luedtke J. Am. Chem. Soc. 1983 105 5030. 23 H. Dahrnane B. Mile H. Morris J. A. Howard and R. Sutcliffe J. Chem. Soc. Chem. Cornmun. 1983 1068. Reaction Mechanisms -Part ( iii) Free-radical Reactions drum and attempts were made to eliminate the obvious possibility that simple pyrolysis was responsible for radical formation. Bilayers of phosphatidylcholine induced differences in the rates of thermal decomposition of the diastereomers of (4).24 The meso-form decomposed six times faster than the (*)-diastereomer in aqueous emulsions of the multilamellar vesicles. Thus the bilayer host had a significant effect on the orientations of the diastereomeric compounds and/or their transition states for decomposition.It was concluded that such bilayers might be used to induce diastereomeric selectivity in synthesis. N It N Niki et al. showed that vitamins C and E could each inhibit the oxidation of methyl linoleate in t-butyl alcohol-methanol solvent.25 The reduction in the oxidation rate during the inhibition period was greater for vitamin E than C presumably because the former acted as a more effective peroxy radical scavenger. However when the two were used in combination the reduction in the rate was the same as that observed for E alone but the length of the induction period reflected the total vitamin concentration. These observations led to the conclusion that vitamin C was regenerating E during the induction period until the two were completely consumed.Similar conclusions were reached by Barclay et a1.,26who carried out a detailed kinetic study of the initiated oxidation of methyl linoleate localized in sodium dodecylsulphate micelles a system which represents a primitive model for biological membrane oxidation. Vitamin E located within the micelles functioned as an effective inhibitor. Again there was a synergistic effect when vitamin C was added even though the latter was localized in the aqueous phase. It was argued that the phenolic function of vitamin E could easily approach the water-impregnated periphery of the micelle where reduction by vitamin C could take place. Reaction of methylthiyl radicals with Z-methyl-4,5-dihydrofurantook place by abstraction of the allylic hydrogen or by addition to the double bond abstraction being favoured at higher temperature^.^' Whereas the radical derived by abstraction could be detected by e.s.r.that derived by addition could not. Instead a rearranged form of the radical was observed which was thought to have resulted from a 1,4 hydrogen migration (Scheme 11). 24 R. C. Petter J. C. Mitchell W. J. Brittain T. J. Mclntosh and N. A. Porter J. Am. Chem. SOC.,1983 105 5700. 25 E. Niki T. Saito and Y. Kamiya Chem. Lett. 1983 631. 26 L. R. C. Barclay S. J. Locke and J. M. MacNeil Can. J. Chem. 1983 61 1288. 27 L. Lunazzi G. Placucci and L. Grossi Tetrahedron 1983 39 159. D. Griller +Q Me Scheme 11 3 Structure There has been a good deal of interest in the structure of boron-containing radicals.The triborane(7) radical anion was detected by e.s.r. spectroscopy when mixtures of di-t-butyl peroxide and Bun4NB3H in a variety of solvents were photolysed in the cavity of the spectrometer.28 The hyperfine splittings were a(l ''B) 7.89 a(2l'B) 2.91 a( 1H) 10.79 a(2H) 11.60 and a(4H) 2.61 G at -45 "C and were consistent with structure (5). This anion was found to be very much less reactive than H3BT as was expected for a delocalized species of higher ionizatibn potential.28 .~~ Symons et ~1 obtained the powder spectrum due to .BH4 on y-irradiation of sodium borohydride or b~rodeuteride.~~ The hydrogen atoms were found to be equivalent in pairs. MNDO-UHF calculations were consistent with these findings and gave the optimized geometry (6).H H \B<,' 58.3"\-/ /-%. B%)I 14.50 H 122.0'H Reaction of silver and gold atoms with acetylene and phenylacetylene at 77 K gave e.s.r. spectra due to metal substituted vinyl and styryl radicals re~pectively.~' The hyperfine splittings were not dissimilar to those observed for the hydrocarbon analogues and the atoms were therefore thought to be a-bonded to carbon. However copper formed mono- and bis-.rr-complexes with acetylene but was a-bonded to phenylacetylene. The e.s.r. spectrum of the tetrafluoroethylene radical cation was obtained by y-irradiation of a solid solution of FCCI containing tetrafluoroethylene and was found to have four equivalent fluorine at~rns.~' The result suggested that the cation had a planar structure which was consistent with INDO calculations.An additional small fluorine hyperfine splitting was thought to be due to a matrix fluorine atom. 28 J. R. M. Giles V. P. J. Marti and B. P. Roberts J. Cbem. SOC.,Cbem. Commun. 1983 696. 29 M. C. R. Symons T. Chen and C. Glidewell J. Cbem. SOC.,Chem. Commun. 1983 326. 30 J. H. B. Chenier J. A. Howard B. Mile and R. Sutcliffe J. Am. Chem. SOC.,1983 105 788. 3' A. Hasegawa and M. C. R.Symorts J. Cbem. SOC.,Faraday Trans. I 1983 79 93. Reaction Mechanisms -Part (iii) Free-radical Reactions 95 By contrast calculation showed that the tetrafluoroethylene radical anion had a chair structure.32 Careful analysis of the e.s.r. spectra of this radical which again showed four equivalent fluorine atoms at -170 "C,was now found to be consistent with the chair structure on the assumption that the radical can rotate about an axis perpendicular to the plane containing the four fluorine atoms.Readers are advised to study these papers in detail since the arguments are more compelling than this summary indicates. N,N'-Bis(ary1thio)benzamidinyl radicals (7) were found to be a new class of persistent nitrogen-centred radicals.33 They were generated by oxidation of the corresponding benzamidines using lead dioxide and had half-lives of hours even in the presence of oxygen. Their persistence was thought to be associated with their delocalized electronic structures rather than with steric protection of the radical centre. N-Alkyl-N-(a1kylthio)aminylradicals were not persistent except for the case when the alkyl was t-b~tyl.~~ However even in this instance reaction with oxygen was rapid suggesting that the longevity of the radical under oxygen-free conditions was due to steric protection of the radical centre.A number of copper( 11) dithiolate complexes were characterized by e.s.r. spec- troscopy and may be of particular biochemical imp~rtance.~' For example d-penicillamine is effective in treating physiological disorders associated with excessive copper levels and was indeed found to produce a planar complex (8) with a half-life of several minutes. The e.s.r. spectra of the trans and cis radical ions (9) and (10) were obtained by photolysis of a-complexes formed between dimerized t-butylmethylacetylene and aluminium chloride.36 The radicals had slightly different hyperfine splittings a(2Me) 9.00G a(2Bu') 0.20G (9) and a(2Me) 8.00G a(2Bu') 0.24G (lo) at -80°C in methylene chloride.It was possible to interpret these differences in terms of breaking the degeneracy of molecular orbitals in the trans-isomer by the more powerful electron release of the t-butyl groups. However it was concluded that such arguments were unreliable because of the likelihood that steric perturbations could have been the origin of these effects. HU' Bu' Bu' (9) 32 A. Hasegawa and M. C. R. Symons J. Chem. SOC.,Faraday Trans. I 1983 79 1565. 33 Y. Miura T. Kunishi and M. Kinoshita Chem. Lett. 1983 885. 34 Y. Miura H. Asada M. Kinoshita and K.Ohta J. Phys. Chem. 1983 87 3450. 35 F. J. Davis B. C. Gilbert R. 0.C. Norman and M. C. R. Symons J. Chem. SOC.,Perkin Trans. 2 1983 1763. 36 J. L. Courtneidge A. G. Davies and J. Lusztyk,J. Chem. Soc. Chem. Commun. 1983 893. 96 D. Griller Photolysis of pentamethylcyclopentadienes with 13C in the methyl groups or in the ring gave the corresponding pentamethylcyclopentadienyl radical^.^' The 13C hyperfine splittings in hexane solvent at -25 "C were a(I3C,j 2.68 G and a(13Cp) 3.35 G demonstrating that they are .rr-delocalized radicals which can be treated by rr-electron theory. 4 Kinetics Rate constants for the reactions of radicals with oxygen were measured in solution using a flash photolysis technique in which the lifetimes of the radicals were monitored as a function of the oxygen c~ncentration.~~ For hydrocarbon radicals the reactions were approximately diffusion-limited bearing in mind that only one- third of the encounters between the doublet radicals and triplet oxygen should lead to doublet products.The rate constants k at 25 "C were therefore relatively insensi- tive to structure e.g. k(t-butyl) = 4.9 x lo9 (in cyclohexanej and k(cyclohexadieny1) = 1.64 x lo9 dm3 mol-' s-' (in benzene). The rate constant for tri-n-butylstannyl radicals was unusually high (7.5 x lo9 dm3 mol-' s-lj a fact that was attributed to relaxation of the spin selection criteria as a result of a heavy-atom effect. Interestingly diphenylaminyl did not react with oxygen in the time-scale of these experiments and the rate constant for its reaction was therefore less than lo7dm3 mol-' s-I.Phenyl radical kinetics were also investigated by flash photolysis technique^.^^ Representative rate constants for its reactions are chlorobenzene I .2 x lo6 carbon tetrachloride 7.8 x lo6,and methyl methacrylate 1.8 x lo8dm3 mol-' s-I. Since the reaction products were not investigated these rate constants represent global rate constants for all possible modes and sites of attack at the substrates. No evidence could be found for the participation of benzoyloxyl radicals when phenyl was generated by photolysis of benzoyl peroxide and the lifetime of the former was concluded to be less than lop8s. This contrasts with the trapping of benzoyloxyl in thermally initiated reactions of benzoyl peroxide and could be taken as support for the existence of two electronic states of the radical (see above).Two groups found similar results for the decarbonylation of the phenylacetyl radical and obtained A factors of ca. 10' 13 s-' and activation energies of 26 kJ mol-' which were little affected by ~olvent.~~~~~ Similarly the @-scission of cumyloxy was described by an A factor of 10'2.4 s-l and an activation energy of 36 kJ mol-' by combination of data obtained in chlorobenzene and cumene as solvents.42 Tri-n-bytylgermanium hydride was found to react with primary alkyl radicals about 20 times more slowly than the corresponding tin hydride at 25"C.43 This property may make it a useful alternative to tin hydride for competitive scavenging of alkyl radicals formed in slow rearrangement reactions.37 A. G. Davies E. Lusztyk J. Lusztyk V. P. J. Marti R. J. H. Clark and M. J. Stead J. Chem. SOC.,Perkin Trans. 2 1983 669. 38 B. Maillard K. U. Ingold and J. C. Scaiano J. Am. Chem. SOC.,1983 105 5095. 39 J. C. Scaiano and L. C. Stewart J. Am. Chem. SOC.,1983 105 3609. 40 L. Lunazzi K. U. Ingold and J. C. Scaiano J. Phys. Chem. 1983 87 529. 41 N. J. Turro 1. R.Gould and B. H. Baretz J. Phys. Chem. 1983 87 531. 42 A. BaignCe J. A. Howard J. C. Scaiano and L. C. Stewart J. Am. Chem. SOC.,1983 105 6120. 43 J. Lusztyk B. Maillard D. A. Lindsay and K. U. Ingold J. Am. Chem. SOC.,1983 105 3578. Reaction Mechanisms -Part ( iii) Free-radical Reactions 97 Livingston and Zeldes investigated the pyrolysis of benzyl ether using a high-pressure-high-temperature e.s.r.cell.44They found the reaction described in Scheme 12 to be an essential feature of the mechanism and determined its A factor ( s-') and activation energy (65.1 kJ mol-'). The technique itself is a considerable innova-tion for e.s.r. spectroscopy and will hopefully be adopted by other investigators. PhCHOCH,Ph -+ PhCHO + CH2Ph Scheme 12 Rate constants were measured for the self-reactions of substituted ben~yl,~~ iso-propylol,46 and a series of 'capto-dative' stabilized radicals.47 Those for the benzyl radicals45were not perfectly described by the von Schmolochowski equation an effect which was ascribed to the unreliability of estimated reaction diameters for bulky radicals.As expected rate constants for isopropyl01~~ were much lower than predicted by theory in solvents capable of hydrogen bonding to the radical. Not surprisingly the capto-dative stabilized radicals ( 1 1)-( 13) all underwent self-reaction at the diffusion-controlled rate.47Obviously the stabilization is insufficient to weight the radical-dimer equilibrium in favour of the radical at ambient tem-peratures. Similarly the stabilized radicals cyanomethyl 2-cyano-2-propy1 syn-and anti-1-cyanoallyl radicals were not persistent and all underwent diff usion-controlled self-reaction?' BU~-O-CH-CN BU~-S-CH-CN MeO-CH-C0,Me (1 1) (12) (13) Burton et al. investigated the properties of a number of phenols e.g. (14)-( 17) which were structurally related to a-tocopherol (14) to see whether the latter had a fully optimized structure as a chain-breaking anti~xidant.~~ Of these (17) was substantially more reactive towards peroxyls than a-tocopherol itself.The results were found to be related to polar and conformational effects manifest in the e.s.r. spectra of the phenoxyl radicals derived from these inhibit01-s.~' Me Relative n R' R2 X reactivity 0 1.o Ho@yf~ (15) C,,H,, (14) 2 Me C(0)OMe 0 0.56 Me X (16) 2 H H NC(0)Me 0.04 I Me (17) 1 H Me 0 1.66 5 Thermochemistry Heats of formation and ionization potentials for a series of a-aminoalkyl radicals R' NCR2 were determined from measurements of appearance energies for the 44 R. Livingston and H.Zeldes J. Phys. Chem. 1983 87 1086. 45 R. F. C. Claridge and H. Fischer J. Phys. Chem. 1983 87 1960. 46 M. Lehni and H. Fischer Int. J. Chem. Kiner. 1983 15 733. 47 H.-G. Korth R. Sustmann R. Merinyi and H. G. Viehe J. Chem. SOC.,Perkin Trans. 2 1983 67. 48 H.-G. Korth P. Lommes W. Sicking and R. Sustmann Int. J. Chem. Kinet. 1983 15 267. 49 G. W. Burton L. Hughes and K. U. Ingold J. Am. Chem. SOC.,1983 105 5950. 50 T. Doba G. W. Burton and K. U. Ingold J. Am. Chem. SOC.,1983 105 6505. 98 D. Griller fragmentations of a series of ethylenediamines.' ' Stabilization energies increased and ionization potentials decreased with increasing C-or N-alkylation. For example when R' = Me R2 = H the methane-based stabilization energy was 84 kJ mol-' and the ionization potential was 5.7 eV as compared with 7-10 eV for most alkyl radicals.E.s.r. spectroscopy was used in several instances to determine thermochemical properties. Jenkins and Perkins used such measurements to determine the 0-H bond strengths in a series of N-t-butylhydroxamic acids with respect to that of (1 8) (Scheme 13).52For example those for (19) and (20) were found to be 334 and 3 15 kJ mol-' respectively. In general these bond strengths were greater than those for dialkyl- nitroxides and increased with electron demand in the acyl group. 0 0 Bu' ROX-N'''' \0. + (18)-H \OH + I tk (19) R = NO 0. (20) R = OMe (18) Scheme 13 Schlosser and Steenken found that (CF3S) N-N(SCF,) underwent reversible homolytic cleavage in perhalogeno-alkane solvents in the temperature range 250-315 K to give the corresponding aminyl radicals which were easily detected by e.~.r.~~ The bond dissociation energy was found to be only 32 f2 kJ mol-'.This facile homolysis was explained in terms of both steric and electronic factors. The rotational barriers for aminopropynyl and aminocyanomethyl were deter- mined from the temperature dependencies of their e.s.r. spectra which were in turn used to calculate their methane-based stabilization energies (107 and 95 kJ mol-' re~pectively).~~ These were equal within experimental error to the sum of the stabilization energies associated with each part of the radical e.g. E (aminocyanomethyl) = E (aminomethyl) + E (cyanomethyl). If the popular 'capto-dative effect' has any substance if should be revealed by tests of this kind.That is the stabilizing effect of the combination of substituents should be greater than the sum of the parts. However such experimental verifications have not yet been adequately demonstrated. A recent claim55 for an additional capto-dative stabilization of 16 kJ mol-' for CH(0Me)CN is almost certainly insignificant when compared with the experimental errors and assumptions built into the experiment although the approach represented a serious test of the capto-dative idea. While theory supports the capto-dative concept it is arguable that the present widespread use of the term lacks definitive experimental support. Two st~dies~~,~' dealt with bond dissociation energies in fluoroalkanes.A gas-phase bromine buffer system was used to determine D(i-C3F7-H) as 433.3 f 5' T. J. Burkey A. L. Castelhano D. Griller and F. P. Lossing J. Am. Chem. Soc. 1983 105 4701. 52 T. C. Jenkins and M. J. Perkins J. Chem. SOC.,Perkin Trans. 2 1983 717. 53 K. Schlosser and S. Steenken J. Am. Chem. Soc. 1983 105 1504. 54 D. Griller D. C. Nonhebel and J. C. Walton J. Chem. Soc. Perkin Trans. 2 1983 1373. 55 R. Louw and J. J. Bunk Recl. J. R. Nerh. Chem. Soc. 1983 102 119. 56 J. P. Martin and G. Paraskevopoulos Can. J. Chem. 1983 61 861. 57 B. S. Evans 1. Weeks and E. Whittle J. Chem. Soc. Faraday Trans. 1 1983 79 1471. Reaction Mechanisms -Part (iii) Free-radical Reactions 2.4 kJ mol-I. However (CF&C-H was found to be particularly unreactive towards bromine atom attack so that the technique could not be applied.Finally an elegant of the thermolysis of tetrasubstituted succinonitriles led to a ‘resonance energy’ of 21 kJ mol-’ for tertiary a-cyanoalkyl radicals when careful account was taken of the steric and electronic factors involved in the dissociation process. W. Barbe H.-D. Beckhaus and C. Ruchardt Chem. Ber. 1983 116 1042.